Work in the past year on the virus plaque assay has centered on two main problems defining the cell population of murine T-cells detected by this assay, and ascertaining whether in human lymphocytes it measures activated T- or B-cells or both populations. In the mouse in secondary MLC, 70% of the virus plaque forming cells are of the Ly 1 phenotype, 30% being of the Ly 2,3 phenotype. Under conditions in which proliferating T-cells were inhibited by cytosine arabinoside, comparable numbers of V-PFC to those found in the Ly 2,3 population were detected, as were equal numbers of killer cells in the drug treated and untreated populations. These studies indicated a parallelism between the proliferation-independent virus plaque forming cells and cytotoxic T-lymphocytes. Preinfection of resting lymphocytes by VSV followed by activation in MLC abolished the cytotoxic T response thereby demonstrating formally that killer T-cells were able to produce virus. The appearance and disappearance with time of this antigen-sensitive non-dividing cells population was studied, and found to be a "prekiller" population, capable of differentiating into a cytotoxic T-lymphocyte in the absence of cell proliferation. Preliminary studies on human tonsillar lymphocytes using separated cell populations and selective stimulation by PHA or the Cowan I strain of staphylococcus are consistent with the interpretation that the virus plaque assay applied to human cells detects activated T- but not activated B-lymphocytes.